The Dynamics of War and Peace in the Aftermath of Natural Disasters: Duration, Sequences, and Political Context

Climate change increases frequency of natural disasters, e.g. storms or floods. Beyond their immediate physical impact, disasters may also trigger significant political consequences such as social unrest or civil wars. These political fallouts often prove to be even more devastating in terms of people affected than the disaster’s immediate physical impacts. In the past the relation between natural disasters and civil wars has attracted considerable attention. Previous studies in particular have investigated whether natural disasters might trigger civil wars or end them and demonstrated that and how the consequences depend on the respective political setting in which they occur. However, many of these studies focus on whether a particular, single event or outcome such as war or peace results from a natural disaster. Yet, very few studies have investigated the dynamics of civil wars following natural disasters. We attempt to close this gap and therefore focus on the series of events after a natural disaster. Hence, we analyze whether and when a natural disaster ignites a prolonged phase of civil wars or repeated circles of war and peaceful periods. Specifically, we distinguish three phases of civil wars, onset, prolonged conflict, and pacification, and investigate a) the duration of the respective phase after a natural disaster occurs and b) the specific sequences of phases that follow a natural disaster. We argue that both the duration of a phase and the specific sequence depend on the particularities of the disaster and on the particularities of the society where it occurs. More precisely, we presume that it matters whether a disaster has immediate consequences, e.g. floods or storms, or delayed consequences, e.g. droughts. We expect stronger effects, i.e., a shorter time period to result in civil wars and longer sequences of civil wars, for the former. We also hypothesize that disasters will differently affect the propensity for civil wars to start or end depending on political factors, e.g. ethnic fractionalization or stable and unstable democracies. For example, we assume longer periods of war and a higher propensity to repeat cycles of war and (short) peaceful periods with higher ethnic fractionalization and within less developed democracies. We generate a detailed world-wide monthly dataset from the EM-DAT, the International Disaster Database that covers natural disasters, the UCDP/PRIO Armed Conflict Dataset that covers civil wars, the Quality of Government Institute Standard dataset (QoG), and the Varieties of Democracy dataset (V-Dem) to obtain extensive information on internal conflicts, national political and economic data, as well as natural disasters from 1946-2024, i.e., the entire period after WWII. To investigate the duration of the respective phases we employ event-history analysis, i.e., survival curves and Cox regression. To investigate the specific sequence following a specific disaster we will employ sequence analysis.